use of an electronic nose to evaluate disease activity in ulcerative colitis
使用电子鼻评估溃疡性结肠炎的疾病活动
zibing j. woodward, kevin piro, sarah lee, emile latour, jodi lapidus, suni wilson,
david lieberman, judith collins, kian keyashian, nir modiano
background and aimsassessment of disease activity in inflammatory bowel disease (ibd) reflects a major challenge in clinical practice, relying on invasive endoscopic evaluation, expensive imaging tests and biomarkers that are difficult to obtain in a timely fashion. cyranose 320 is an electronic nose that senses volatile organic compounds; it has been previously used to distinguish between malignancies, inflammatory states and infections. in this proof of concept study, we analyzed the urine headspace of active and inactive ulcerative colitis (auc, iauc) patients (pts) as well as non-ibd controls with cyranose in an effort to differentiate between groups to enable real-time disease assessment. methods urine samples were collected from consecutive uc pts seen in ibd clinic and during inpatient admissions; control subjects were those undergoing outpatient upper endoscopy for nonmalignant indications. auc pts were those with moderate activity classified as a partial mayo score of ³ 5 or full mayo score of ³ 6, inactive uc were those with a partial mayo score of 0 or 1 and endoscopic mayo score of 0 or 1. the urine samples were stored in a-70°c freezer until the analysis was carried out. at that time, samples were thawed in a 40°c water bath and handled in a tissue culture laminar flow hood; urine headspace was analyzed by cyranose according to machine protocol. the raw data collected from the 32 proprietary sensors were then analyzed using one-way analysis of variance (anova), canonical discriminant analysis (cda) and logistic regression.
炎症性肠病(ibd)疾病活动评估反映了临床实践中的一个主要挑战，依赖于侵入性内镜评估、昂贵的成像测试和难以及时获得的生物标志物。cyranose 320是一种电子鼻，能感知挥发性有机化合物；它以前被用来区分恶性肿瘤、炎症状态和感染。在这项概念验证研究中，我们分析了活动性和非活动性溃疡性结肠炎(auc，iauc)患者(pts)和非ibd对照者使用cyranose 320电子鼻尿液顶空分析，以区分各组，以便进行实时疾病评估。方法从ibd门诊及住院期间连续出现的uc患者中采集尿液标本，对照组为非恶性适应症门诊上内镜检查。auc评分为中度活动组，部分梅奥评分为3.5分或全部梅奥评分为3.6分；非活动组，部分梅奥评分为0或1分，内镜梅奥评分为0或1分。尿液样本储存在-70°c的冷冻室中，直到进行分析。当时，样品在40°c水浴中解冻，并在组织培养层流罩中处理；根据机器方案，用cyranose尿液顶空分析。然后采用方差分析(anova)、标准判别分析(cda)和逻辑回归对32个专有传感器的原始数据进行分析。
results a total of 41 samples were analyzed: 12 pts with auc, 14 with iauc and 15 controls. one-way anova showed a statistically significant difference between group means for sensors 5-9, 11, 23, 29 and 31. false discovery rate was calculated to be < 1.0% for sensors 6 and 31, indicating a significant difference between group means using these sensors. the importance of sensors 6 and 31 was also shown using random forest, a recursive partitioning approach. canonical discriminant analysis shows a significant separation between the three groups (figure 1). logistic regressions using only sensors 6 and 31 were able to differentiate between pts with auc and iauc with good accuracy (area under the receiver operating characteristic curve - auroc = 0.80). similarly, these sensors were also able to distinguish between auc and control pts (auroc = 0.85), iauc and controls (auroc = 0.86) as well as uc pts as a whole and controls (auroc = 0.78; figure 2).
共分析41个样本：12个auc患者，14个iauc患者和15个对照组。单因素方差分析显示，传感器5-9、11、23、29和31的组间平均值存在统计学显著差异。对于传感器6和31，错误发现率计算为<1.0%，表明使用这些传感器的组间平均值存在显著差异。传感器6和31的重要性是还显示了使用随机林的递归分区方法。典型判别分析显示三组之间存在显著的分离(图1)。仅使用传感器6和31的逻辑回归能够很好地区分具有auc和iauc的pts(接收器操作特性曲线下的面积-auroc=0.80)。同样，这些传感器还能够区分auc和对照组(auroc=0.85)、iauc和对照组(auroc=0.86)以及uc整体和对照组(auroc=0.78；图2)。
conclusion cyranose 320 is a novel technology for analyzing volatile organic compound profiles. this study shows it can be used to distinguish between pts with active and inactive uc as well as those without ibd. cyranose may become a useful tool for point-of-care assessment of ibd disease activity. larger studies are required to validate these findings.
cyranose 320是一种分析挥发性有机化合物分布的新技术。这项研究表明，它可以用来区分有活动和不活动uc的pts以及没有ibd的pts。需要更多的研究来验证cyranose 成为护理点评估ibd疾病活动新工具